Insight and Open Science

advertisement
The Insight Toolkit
Building
Open Science
Communities
Luis Ibáñez
Kitware, Inc.
Insight Software Consortium
What is ITK ?

Software Library

Written in C++

Using Generic Programming

Extreme Programming

Open Source with Free license
What is Free Software ?
“Free software” is a matter
of liberty, not price.
To understand the concept,
you should think of “FREE” as
in “FREE SPEECH”,
not as in “FREE BEER”.
http://www.gnu.org/philosophy/free-sw.html
Open Source Software
versus Free Software
“Open source is a development methodology”
“Free software is a social movement”
“For the Open Source movement, non-free software is
a suboptimal solution.”
“For the Free Software movement, non-free software
is a social problem and free software is the solution.”
http://www.gnu.org/philosophy/free-software-for-freedom.html
Interdit d’Interdire
ITK is Free
as in
“Free Speech”
and…
as in
“Free Beer”
Well…
free for the users…
ITK Sponsors
The National
Institute for Dental and
Craniofacial Research
The National Institute of Neurological
Disorders and Stroke
The National
Science
Foundation
ITK Developers
Insight Software Consortium
How did ITK start ?
Visible
Human
Project
Acquisition
MR, CT &
Cryogenic
Distribution
but,
Data was not Enough…
Software was needed
for analyzing the data
Use my library…
I have
one, but…
I have
one, but…
I have
one, but…
I have
one, but…
I have
one, but…
Developing Software
for Research
is an intrinsically
Ungrateful
business
Data
Driving
Problem
Software
Research
Mean
Goal
?
Algorithms
Papers
You don’t get research
credits for:








Implementing algorithms
published by others
Writing Software Documentation
Fixing Bugs
Improving Performance
Preparing Tutorials
Porting to new platforms
Supporting Users
Making software releases
If you are a student
Software will not give
you a degree…
If you are a professor
Software will not give
you a promotion…
Software development
is seen as
not worthy
of a researcher time
Raise your hand
those who can do
Medical Image Processing
without
Software
You do get research
credits for:





Publishing papers
Publishing books
Getting Patents
Getting Funding (Grants, Contracts)
Licensing your Patents
Why is that ?
Time to face the
Truth
Publications
do not
cure Cancer !
Doctors do not prescribe
“reading papers”
as a treatment.
Medical treatment is done with
 Medical
Devices
 Drugs
 Surgical
Procedures
Publications that don’t lead
to one of those treatments
are sterile publications
Really good
research results
are not published…
They get Patented !
With the hope of being used for
 Medical
Devices
 Drugs
 Surgical
Procedures
Why do we care so much
about publishing ?
Publications are a measure
of scientific productivity

They disseminate knowledge

They allow others to reproduce our results

They are validated by the peer-review process
Papers disseminate
knowledge
Information in the 21st Century
Is disseminated
on the Internet
How long it takes to post a
PDF file on the Web ?
At most 1 day
Typically 1 hour
How long it takes to publish
a paper on a Journal ?
At least 1 year
Typically 2 years
How much do you
have to pay for publishing
a paper in a Journal ?
About $500 / paper
How much do you
have to pay
for reading the same paper ?
About $30 / paper
or subscribe for $300 / year
How much it costs to
post a PDF on the Web ?
Certainly less than
$500 + N x $30
Papers allow others to
reproduce the results
Reproducing the Results…

Do you get source code with the paper ?

How long it will take you to rewrite this code ?

Do you get the author’s data ?

How can you get their data ?

Do you get all the parameters they used ?

How can you reproduce results if you don’t
have code, data and parameters ?
And anyways, why do you
want to invest time in reproducing
somebody else’s results…
If you don’t get
any credit for doing it ?
Have you ever seen a paper
in a Medical Image Journal
whose only content is the
reproduction of results from
another paper ?
Have you ever seen a paper
in a Medical Image Journal
whose only content is the
failure to reproduce the results
of another paper ?
If reproducibility is the goal
of publishing…
 You
should post your source code
 You
should post your data
 You
should post your parameters
In the same way that you posted your
PDF file: on the Web.
Research is validated
by the
Peer-Review process
How can a reviewer
validate a paper ?
If we just concluded
that papers are not
reproducible…
What does a reviewer
actually do ?
Emit an opinion based on
his/her expertise
How much time does a reviewer
dedicate to a paper ?

1 hour ?

2 hours ?

6 hours ?
Why not more time ?

Reviewers are volunteers

They don’t get paid for reviewing papers

They don’t get credits for reviewing papers

They have their own papers to write

They have exams to grade

Their own grant applications to submit

They also have families, pets and… a life !
How long does a paper waits on
the reviewer’s desk before he/she
finds time for reviewing it ?
 Six
6
weeks ?
months ?
How many reviewers typically
judge your paper ?
 Minimum
Two
 Typically
Three
 Exceptionally
Four
 Why
not more ?
 Why
only one time ?
Why do we really
want to publish ?
Because we need
to have publications
in our CV
“Publish or Perish”
Who invented this ?
and Why ?
“Publish or Perish”
Was invented by those who
needed to evaluate
researcher’s productivity.
“Publish or Perish”
Empowers those who read
your CV to grade you by
simply counting lines in the
“Publications” section.
“Publish or Perish”
The group of best educated
people in the world has been
alienated with a simple trick
Who are you
working for ?
Who really pays
your salary ?
Who pays for Research ?
Public
Medical Device
Manufacturers
Hospitals
& Doctors
Researchers
Pharmaceutical
Companies
What do your owe to those
who pay your salary ?
Competition with other
researchers ?
or
Collaboration with other
researchers ?
How to collaborate ?





Creating public repositories for source code
Creating public image databases
Posting parameters on the web
Creating forums for hosting positive
discussions online
Validating other’s methods and suggesting
improvements.
The Insight Journal Solution
Open Source
Open Science
Insight
Journal
Agile Programming
Agile Publishing
The Dark Ages are Over…
Embrace Open Science !
Download